def test_get_split():
with pytest.raises(Exception):
t = ToolBox(X=X, y=y, query_type='AllLabels', saving_path=None)
a, b, c, d = t.get_split()
tb.split_AL(test_ratio=0.3, initial_label_rate=0.1, split_count=split_count)
a, b, c, d = tb.get_split()
assert (check_one_to_one_correspondence(a, b, c, d))
def test_init():
with pytest.raises(ValueError):
ToolBox(y[0:5], X, query_type='AllLabels', saving_path=None)
with pytest.raises(NotImplementedError):
ToolBox(y, X, query_type='AllLabel', saving_path=None)
with pytest.raises(Exception):
ToolBox(y, x=None, query_type='Features', saving_path=None)
n_features=20,
n_informative=2,
n_redundant=2,
n_repeated=0,
n_classes=2,
n_clusters_per_class=2,
weights=None,
flip_y=0.01,
class_sep=1.0,
hypercube=True,
shift=0.0,
scale=1.0,
shuffle=True,
random_state=None)
split_count = 5
acebox = ToolBox(X=X, y=y, query_type='AllLabels', saving_path=None)
# split data
acebox.split_AL(test_ratio=0.3,
initial_label_rate=0.1,
split_count=split_count)
# use the default Logistic Regression classifier
model = acebox.get_default_model()
# query 50 times
stopping_criterion = acebox.get_stopping_criterion('num_of_queries', 50)
# use pre-defined strategy, The data matrix is a reference which will not use additional memory
randomStrategy = QueryRandom()
uncertainStrategy = QueryInstanceUncertainty(X, y)
import os
import numpy as np
from sklearn.datasets import load_iris
from acepy.experiment import State, StateIO
from acepy.toolbox import ToolBox
X, y = load_iris(return_X_y=True)
split_count = 5
cur_path = os.path.abspath('.')
toolbox = ToolBox(X=X, y=y, query_type='AllLabels', saving_path=cur_path)
# split data
toolbox.split_AL(test_ratio=0.3, initial_label_rate=0.1, split_count=split_count)
train_ind, test_ind, L_ind, U_ind = toolbox.get_split(round=0)
# -------Initialize StateIO----------
saver = StateIO(round=0, train_idx=train_ind, test_idx=test_ind, init_L=L_ind, init_U=U_ind, saving_path='.')
# or by using toolbox
# saver = toolbox.get_stateio(round=0)
saver.init_L.difference_update([0, 1, 2])
saver.init_U.update([0, 1, 2])
# -------Basic operations------------
st1_batch1 = State(select_index=[1], performance=0.89)
my_value = 'my_entry_info'
st1_batch1.add_element(key='my_entry', value=my_value)
st1_batch2 = State(select_index=[0, 1], performance=0.89)
st2_batch1 = State(select_index=[0], performance=0.89)
st3_batch1 = State(select_index=[2], performance=0.89)
from __future__ import division
import pytest
from sklearn.datasets import load_iris
from acepy.toolbox import ToolBox
from acepy.utils.misc import check_one_to_one_correspondence
X, y = load_iris(return_X_y=True)
split_count = 5
tb = ToolBox(X=X, y=y, query_type='AllLabels', saving_path=None)
def test_init():
with pytest.raises(ValueError):
ToolBox(y[0:5], X, query_type='AllLabels', saving_path=None)
with pytest.raises(NotImplementedError):
ToolBox(y, X, query_type='AllLabel', saving_path=None)
with pytest.raises(Exception):
ToolBox(y, x=None, query_type='Features', saving_path=None)
# with pytest.raises(TypeError):
# ToolBox(X=X, y=y, query_type='AllLabels', saving_path='asdfasf')
def test_al_split():
train_idx, test_idx, Lind, Uind = tb.split_AL(test_ratio=0.3, initial_label_rate=0.1, split_count=split_count)
assert (check_one_to_one_correspondence(train_idx, test_idx, Lind, Uind))
def test_get_split():
with pytest.raises(Exception):
data_root = 'C:\\Code\\AAAI19_exp\\final_exp\\benchmarks_keel.mat'
datasets = scio.loadmat(data_root)
dataname = 'clean1'
data = datasets[dataname]
data = data[0][0]
# print(type(data))
# print(len(data))
# print(data)
X = data[0]
y = data[1].flatten()
# X, y = load_digits(return_X_y=True)
# X, y = make_classification(n_samples=150, n_features=20, n_informative=2, n_redundant=2,
# n_repeated=0, n_classes=2, n_clusters_per_class=2, weights=None, flip_y=0.15, class_sep=1.0,
# hypercube=True, shift=0.0, scale=1.0, shuffle=True, random_state=None)
acebox = ToolBox(X=X, y=y, query_type='AllLabels', saving_path='.')
# Split data
acebox.split_AL(test_ratio=0.3, initial_label_rate=0.1, split_count=10)
# Use the default Logistic Regression classifier
model = acebox.get_default_model()
# The cost budget is 50 times querying
stopping_criterion = acebox.get_stopping_criterion('num_of_queries', 50)
oracle1 = Oracle(labels=[1] * len(y))
oracle2 = Oracle(labels=[-1] * len(y))
oracles = Oracles()
oracles.add_oracle(oracle_name='Tom', oracle_object=oracle1)
oracles.add_oracle(oracle_name='Amy', oracle_object=oracle2)
from sklearn.datasets import load_iris
from acepy.toolbox import ToolBox
X, y = load_iris(return_X_y=True)
acebox = ToolBox(X=X, y=y, query_type='AllLabels', saving_path='.')
acebox.split_AL(test_ratio=0.3, initial_label_rate=0.1, split_count=10)
model = acebox.get_default_model()
train_idx, test_idx, Lind, Uind = acebox.get_split(0)
# -------------Initialize---------------
# initilize a strategy object by ToolBox
QBCStrategy = acebox.get_query_strategy(strategy_name='QueryInstanceQBC')
# import the acepy.query_strategy directly
from acepy.query_strategy import QueryInstanceQBC, QueryInstanceUncertainty
uncertainStrategy = QueryInstanceUncertainty(X, y, measure='entropy')
# --------------Select----------------
# select the unlabeled data to query
model.fit(X[Lind.index], y[Lind.index])
select_ind = uncertainStrategy.select(Lind, Lind, batch_size=1, model=model)
print(select_ind)
# Use the default logistic regression model to choose the instances
select_ind = uncertainStrategy.select(Lind, Uind, batch_size=1, model=None)
# Use select_by_prediction_mat() by providing the probabilistic prediction matrix
prob_mat = model.predict_proba(X[Uind.index])
select_ind = QBCStrategy.select_by_prediction_mat(unlabel_index=Uind,
predict=prob_mat,
batch_size=1)
print(select_ind)
data_root = 'C:\\Code\\AAAI19_exp\\final_exp\\benchmarks_keel.mat' datasets = scio.loadmat(data_root) dataname = 'clean1' data = datasets[dataname] data = data[0][0] # print(type(data)) # print(len(data)) # print(data) X = data[0] y = data[1].flatten() # X, y = load_digits(return_X_y=True) # X, y = make_classification(n_samples=150, n_features=20, n_informative=2, n_redundant=2, # n_repeated=0, n_classes=2, n_clusters_per_class=2, weights=None, flip_y=0.15, class_sep=1.0, # hypercube=True, shift=0.0, scale=1.0, shuffle=True, random_state=None) acebox = ToolBox(X=X, y=y, query_type='AllLabels', saving_path='.') # Split data acebox.split_AL(test_ratio=0.3, initial_label_rate=0.1, split_count=10) # train_idx, test_idx, label_ind, unlab_ind = acebox.get_split(round=0) # # bmdr = QueryInstanceBMDR(X, y, kernel='linear') # select = bmdr.select(label_ind, unlab_ind) # print(select) # # spal = QueryInstanceSPAL(X, y, kernel='linear') # select = spal.select(label_ind, unlab_ind) # print(select) # # lal = QueryInstanceLAL(X, y, mode='LAL_iterative', train_slt=False)